Imagine joining a software team today.
You have access to the GitHub repository, Jira tickets, documentation, and Slack. Everything is technically available.
But you still don't know why the architecture looks the way it does, why a particular approach was rejected, what broke during the last release, or which decisions are still unresolved.
So you start asking people.
The problem? The answers are scattered across months of conversations, pull requests, tickets, and documents. The people who know the history are busy building something else.
This isn't just an onboarding problem. It happens every time someone joins a project, changes teams, revisits an old decision, or takes over work from another engineer.
The missing piece in AI-assisted development
AI coding assistants are becoming increasingly capable, but they can only use the context available to them.
Your coding assistant might understand the code in front of it without knowing why your team made a particular decision six months ago.
It might see a bug report without knowing the discussion that led to the current implementation.
It might suggest a solution without knowing that the team already tried and rejected it.
The challenge isn't always making AI smarter. Sometimes, it's giving AI access to the right knowledge.
That's the problem I'm working on with LISA
I'm building LISA to help companies make their existing knowledge useful to employees and AI tools.
The goal is to connect information across company systems and help people:
Find answers without manually searching multiple tools.
Understand the history behind project decisions.
Catch up on ongoing work, open questions, and next steps.
Preserve context across team changes and project handoffs.
Verify answers against the underlying company sources.
The vision goes beyond search. I want company knowledge to remain useful even when the people who originally created it aren't available to explain it.
An example
Imagine a developer asks:
"Why did we choose this authentication flow, what alternatives did we reject, and what still needs to be fixed?"
The answer may require information from a pull request, a Jira ticket, a design document, and an old team discussion.
Today, someone might need to find all four and reconstruct the story manually.
The goal with LISA is to make that process much easier, with answers grounded in the relevant sources and access controlled by the company.
Why I'm building this
I believe companies will increasingly use AI not just to generate code or write documents, but to help people understand and work with everything their organizations already know.
For that to work, company knowledge needs to be accessible, contextual, verifiable, and controlled by the organization.
That's what I'm working toward with LISA.
I'm building in public and would love feedback from developers, engineering managers, and founders.
How does your team preserve the reasoning behind technical decisions?
When a new developer joins your project, can they discover why things work the way they do, or do they have to ask the people who were there?
I'd love to hear how other teams handle this.
Top comments (0)